연구 분야: Networking
학회: SN Computer Science
Edge Computing is now emerging as an essential component of network architectures for guaranteeing short response times to IoT applications that are increasingly demanding in terms of processing power and latency. However, the multitude of IoT applications and the existence of heterogeneous Edge architectures means that we need to define an effective placement solution to determine, in real time, how to best distribute (micro)services between the various Edge Computing servers. To do this, in addition to user mobility, which is a point that is almost systematically taken into account, two other elements seem to be decisive for service placement: (a) interdependency between microservices, for example, for regular access to a database, and (b) prioritization between applications, for example to guarantee the smooth running of critical applications. That is why in this article we propose a fuzzy logic-based placement solution that includes these two ideas (interdependence, priority) and also proposes, for the first time in this field, the idea of taking user priority into account for scenarios of low availability of resources and applications with identical levels of criticality. The evaluations carried out, based on a Kubernetes deployment, tend to demonstrate the potential of our solution compared to existing placement work in terms of performance (in particular latency, reliability of placement), energy efficiency and scalability.
| 발행 연도 | 2025년 |
|---|---|
| 인용수 | 0 |
| 출판 국가 | Cyprus, Andorra, France |
| 사이트 | Springer |
| 좋아요 수 | 0 |